Vol. 4 , No. 6, June 1965 NOTES 915

characteristic of the precipitated fluoride, the following sequence of reactions is proposed : For niobium

S b + 4Agr\T03 --f Nb(N03)4 + 4Ag

2Nb + 10Nb(h’o3)4 + lOCH3OH ---f lOCH3ONO + 10NbO(r\!03)3 + Nbz06 f 5H20

NbO(N0s)s + 6NHdF + (SH4)zNbOFs f 3”dOz

For vanadium, the first two steps are as for niobium, followed by

VO(NOa)3 + 5NH4F -+- (NH4)zVOFs f 3NHaNOs

For titanium

Ti + 4AgNOa + Ti(NO3)4 + 4Ag

T i + 2Ti( S 0 3 ) 4 + 2CHaOH -+- 2TiOH( N03)a f 2CH30N0 + TiOz

Ti(N03)4 + 6iKH4F -+- (NH&TiF6 + 4NH4N03

TiOH( N03)3 + 5NH4F --+ ( NHa)?TiFsOH + 3NHaN03

The postulation of NbO(N03)3 as an intermediate is reasonable. A recent paper1’ indicates that the com- pound has been prepared and that it reacts with potas- sium fluoride to give KsNbOF6. We have also shown12 that a bromide, prepared in methanol but not isolated, reacts with methanolic ammonium fluoride to give (NH4)sNbOFs. A similar argument can be applied to the case of vanadium, although the corresponding VO-

The existence of titanium(1V) nitrate in solution is less definite, and the evidence for the formation of the hydroxy nitrate is inferential. The formation of methyl nitrite must be accompanied by solvent oxida- tion or the introduction of oxygen into the titanium species by some oxidative mechanism. I t does not seem reasonable, however, to assume that oxide ion can be present in the fluoride derivative without sub- stantially altering the X-ray diffraction pattern from the (NH&TiF6 type. Hydroxyl can substitute iso- morphously for fluoride without an alteration in the observed powder pattern.

has not yet been isolated.

(11) B. 0. Field and C. J. Hardy, Pvoc. Chem. Soc., 11 (1963). (12) A. E. Baker and H. M. Haendler, I?zoi‘g. Chein., 1, 127 (1962).

CONTRIBUTION FROM THE DEPARTMENT O F CHEMISTRY, SYRACUSE UNIVERSITY, SYRACUSE, NEW YORK

Charge-Transfer Spectra Involving Anions Outside the Coordination Sphere

BY W. A. BAKER, JR., AND MARY G. PHILLIPS

Receiaied December 14, 1964

The existence of charge-transfer bands in the elec- tronic spectra of coordination compounds is now well established. Depending upon the formal charge on

the metal and the nature of the ligands, such bands may involve either the “oxidation” or the “reduction” of the metal ion.’ In most known cases of charge- transfer bands in coordination compounds, the group involved in the transfer is coordinated to the metal.

Examples in which the group involved is not co- ordinated to the metal are less common LinhardL has observed that a very intense ultraviolet band that appears ip solutions of Co(NH3je3+ shifts to lower energies upon the addition of bromide or iodide ion. He attributes this to charge transier between the anion and the complex cation. We wish to report here two series of compounds in which charge transfer involving anions outside the coordination sphere may be re- sponsible for the observed spectra of the solids.

Experimental Materials.-The synthesis of [Cr( bipy),Clzj C1.2H20 r as ac-

complished by the method of Burstall and Nyholm.3 I n this procedure anhydrous chromium( 111) chloride and the required amount of bipyridyl are dissolved in hot methanol and a trace of zinc dust is added. Upon cooling, a green substance precipitates, and after recrystallization from water, brown [Cr( bipy)2C12]- C1.2H20 is obtained. Anal. Calcd. for C J O H ~ ~ N ~ C ~ ~ C ~ . ~ H ~ O : C, 47.40; H , 3.95; N, 11.06. Found: C, 47.79; H , 4.18; K, 10.89.

The salts [Cr(bipy)zC1,]X.2Hz0 ( X = Br-, SZ-) and [Cr- (bipy),Clz]X (X = I-, SCX-) were prepared by dissolving [Cr- (bipy)~CL]C1.2H20 in warm water and adding an aqueous solu- tion of the sodium salt of the appropriate anion. The different salts immediately precipitated. Anal. Calcd. for CnuHleNr- ClpBrCr.2H20: C, 43.55; H , 3.66; N, 10.17. Found: C, 43.75; H , 3.92; N, 10.07. Calcd. for C ~ O H I ~ N ~ C ~ ~ C ~ 2H20: C, 46.76; H , 3.93; N, 19.10. Found: C, 46.9; H, 3.38; N, 19.4. Calcd. for C2~H~6N4C121Cr: C, 42.74; H , 2.87; N, 9.97. Found: C, 42.13; H , 2.88; N, 10.06. Calcd. for C21- HlaN&12SCr: C, 51.1; H, 3.27; N, 14.20. Found: C, 50.91; H, 3.29; N, 14.70.

X-Ray powder data showed the three dihydrates (Cl-, Br-, Na-) to be isomorphous as were the two anhydrous salts ( I - , SCN-). The colors of the salts were distinctly different, being for the chloride, bromide, and azide diff.rent shades of green, for the iodide yellow, and for the thiocyanate light krown.

The preparation of [Cr(phen)zClz] .2H20 followed exactly that of the analogous bipyridyl complex with the substitution of 1 ,lO-phenanthroline for the bipyridyl. The product recovered from the methanol, however, was red in color and analysis in- dicated that the recovered product was approximately the 4- hydrate. Anal. Calcd. for C24H~6S4C13Cr.4H20: C, 48.79; H,4.09; N,9.48.

The salts [Cr(phen)2C12]X.4H20 ( X = Br-, AT3-, SCS- ) and [Cr(phen)tClz] I .2H20 were precipitated from a warm aqu- eous solution of the chloride by the addition of the appropriate sodium salt. Anal. Calcd. for C24Ht&“C12BrCr.4H20: C, 45.43; H , 3.78; N, 8.83. Found: C, 45.74; H, 3.63; N, 8.67. Calcd. for C24H16N1C12Cr~4H20: C, 48.24; H , 4 02; h’, 16.42. Found: C, 48.05; H , 3.84; N, 16.36. Calcd. for C26H16N5C12SCr.4H~0: C, 48.94; H, 3.92; N, 11.42. Found: C, 49.54; H , 3.52; N, 12.12. Calcd. for C24H16N4C1JCr. 2Hz0: C, 44.58; H , 3.10; X, 8.67. Found: C, 44.54; H, 3.35; N, 9.37.

Again the X-ray powder data show that all the 4-hydrates (Br-, N - , SCN-, red C1-) are isomorphous. The colors of the

Found: C,49.42; H,3.59; N,9.30.

(1) See, for example, L. E. Orgel, “An Introduction to Transition Metal

(2) M. Linhard, 2. Elektvochem., 60, 224 (1944). (3) F . H. Burstall and R. S. Nyholm, J . Chem. Soc., 3377 (19Z2).

Chemistry,” John Wiley and Sons, lUew York, N. Y . , 1960, Chapter 6.

916 NOTES Inorganic Chemistry

salts were, for the iodide and azide, orange-brown, for the bromide, maroon, and for the thiocyanate, light tan.

Spectra.-Solution spectra were obtained in a 50-50 methanol- water solution using a Perkin-Elmer Model 4000X spectrophotom- eter and 1-cm. cells. The spectra of the solids were obtained by reflectance using a Beckman DU equipped with a standard re- flectance attachment.

Discussion The reflectance spectra for the two series of com-

pounds [Cr(phen)2C12]X.nHz0 and [Cr(bipy)zClz]X. nH20 are given in Figures 1 and 2 . They consist of

;(cm-') x IO-^

Figure 1.-Reflectance spectra of [Cr(phen)&&]X.nH?O. The ordinate is relative intensity only.

two main features, a rather weak band a t about 18,000 cm.-l and the beginning of a much more intense band a t higher energies.

For both series of compounds, the band at 18,000 cm.-' can be assigned as a d-d transition, specifically 4Azg += 4Tzg. As would be expected, the energy of this band in the phenanthroline series is, as near as can be determined, independent of the anion For the bipyridyl series, however, the iodide and thiocyanate transitions occur a t noticeably higher energies than the other three. It is interesting that these two are an- hydrous while the other three are dihydrates. Why the water of hydration should cause this difference in the d-d band is not clear

The solution spectra for the five compounds in each series are identical, with the low-energy band appearing a t 18,000 em.-' for the bipyridyls and a t 17,800 cm.-l

1.6 17 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5 v(c rn- l ) X I O - ~

Figure 2.-Reflectance spectra of [Cr( bipyj2C12] X . n H 2 0 . The ordinate is relative intensity only.

for the phenanthrolines. Furthermore, the compounds can be repeatedly dissolved and precipitated without changing either the solution or solid spectra. This is cited only as additional evidence that in all cases two chloride ions are coordinated and the variable anion is not.

The portion of the spectrum of interest is the intense absorption beginning a t higher energies. The intensity of the band indicates that i t is due to an allowed transi- tion and i t is clear that the energy depends on the nature of the anion, especially for the compounds in the phen- anthroline series. The order of energies of the bands for both series is the same, I- < N3- < SCN- < Br- < C1-. This order is the same as found by Linhard, et aLj4 for the charge-transfer bands in the series [Cr(NH3)5X]2+, X = C1-, Br-, I-, N3-, and SCN-. All of this suggests that the high-intensity bands ob- served in our compounds are charge transfer in nature and either directly or indirectly involve the anion out- side the coordination sphere.

Charge-transfer bands of this type undoubtedly appear in the spectra of other complexes, but the effect is particularly striking here because the bands extend into the visible region and thus affect the color of the compounds (see Experimental section). The fact that the effect of the anions is so pronounced in these

(4) M. Linhard, H. Sieber-t, and M. Weigel, 2. unoyg. digem. Chent.. 278, 287 (1Q55).

Vol. 4 , No. 6, June 1965 NOTES 917

complexes is probably due to the conjugated nature of the phenanthroline and bipyridyl which in some way leads to a greater electronic interaction with the anions.

Acknowledgment.-The support of the National Science Foundation is gratefully acknowledged.

CONTRIBUTION FROM THE LAWRENCE RADIATION LABORATORY, UXIVERSITY OF CALIFORNIA, LIVERMORE, CALIFORNIA

The Crystal and Molecular Structure of Tetramethylammonium Hexahydrohexaborate'

BY RILEY SCHAEFFER,~ QUINTIN JOHNSON, AND GORDON S. SMITH

Received December 17, 1964

Eberhardt, et al., have explicitly pointed out that a B6Hs2- ion of octahedral symmetry can be well-de- scribed by a closed-shell electronic structure. In less explicit fashion, the treatment of MBs borides by Longuet-Higgins and Roberts leads to a similar con- clusion, since the internal electronic structure is undis- turbed if the external B-B bonds linking octahedra in the boride structure are replaced by external B-H bonds4 Recent isolation by Boone of salts containing the BeHs2- ion makes it possible to confirm this struc- t ~ r e . ~ We have now carried out a complete X-ray structure analysis of tetramethylammonium hexa- hydrohexaborate.

The powder pattern can be indexed on the basis of a face-centered cubic lattice with a = 11.84 f 0.01 A., suggesting a gross structure of the m2x salt based on the antifluorite structure. The calculated density, as- suming four molecules per unit cell, is 0.89 g. ~ m . - ~ compared with 0.88 * 0.02 g. ~ m . - ~ measured by Boone.

Crystals grown from water solution were examined by precession and Weissenberg methods. Systematic ex- tinctions are consistent with space groups F432, F43m, and Fm3m. The intensities of 59 independent reflec- tions were measured with Mo KCY radiation (A 0.7107 8.) on a G.E. XRD-5 diffractometer equipped with a single- crystal orienter.

The structure was solved by noticing that the anti- fluorite structure can accommodate the octahedral

ion a t the origin and the tetrahedral (disre- garding hydrogen) N(CK&+ ion a t f l / d j l /4, i /4 .

Reasonable values for the two parameters which relate to the size of the ions can be guessed from expected bond distances. Preliminary structure factor calcula- tions verified this model.

(1) Studies of Boranes. XVI. For paper XV of this series see J. Dobson

(2) Department of Chemistry, Indiana University, Bloomington, Ind. (3) W. H. Eberhardt, B. L. Crawford, Jr., and W. N. Lipscomb, J . Chem.

(4) H. C. Longuet-Higgins and M. de V. Roberts, PYOC. Roy . SOC. (Lon-

( 5 ) J. Boone, J . A m . Chem. Soc., 86, 5036 (1964).

and R. Schaeffer, Inorg. Chem., 4, 593 (1965).

Phys , 22, 989 (1964).

don), A224, 336 (1954).

From a difference Fourier based on this structure, the boron-bonded hydrogen atom was readily located, but the methyl hydrogens gave the appearance of free rotation. As an approximation to free rotation, half- hydrogen atoms were introduced in 192-fold general positions of space group Fm3m. Least-squares re- finements yielded bond distances as well as temperature factors that were unacceptable. The resolution of the difference Fourier was poor because only a small amount of data were available. If the methyl hydro- gen atoms were fixed, but had a large thermal para- meter, the appearance would probably be one of free rotation ; therefore, two alternate structures in this space group were considered in which hydrogen atoms were fixed. These structures, which have hydrogen atoms located in mirror planes, differ by a 60" rotation of the methyl group. One structure resulted in no improvement over free rotation, but the other struc- ture gave realistic bond distances and temperature fac- tors as well as a substantial improvement in the R factor.

This structure, with anisotropic temperature factors for boron and carbon, was refined by a full-matrix, least-squares program until the parameter shifts were less than one-tenth of the estimated standard devia- tioms The final R factor was 5.3%. As an approxima- tion to a weighting scheme appropriate for diffracto- metric data,' the following was used: W = FO1/' for F, < A , W = A5//"F0-l for Fo > A . Scaled to the struc- ture factor data in Table 11, A is 6.

Final parameters for all atoms are shown in Table I. The notation is that of "International Tables for X- ray Crystallography," Vol. I.* Observed and calcu- lated structure factors are shown in Table I1 and per- tinent bond distances are given in Table 111.

Powder photographs, using chromium radiation (A KCYI 2.28963 f i . ) , of the corresponding cesium salt show a pseudo-primitive cubic cell, a = 4.78 A., due to the overwhelming scattering of cesium. Weak lines re- quire the true cell to be a face-centered cubic lattice with a = 9.561 If: 0.001 f i . Isomorphism with the tetramethylammonium salt would appear likely.

TABLE T FINAL ATOMIC PARAMETERS FOR [N(CHs)&BeHe

24 B ( e ) x = 0.1007 f 0.0008 011 = 0.0071 & 0.0010 0 2 2 0.0088 f 0.0006

(f) x = 0.3219 f 0.0004 32 C 011 = 0.0137 f 0.0005 012 -0.0065 io 0,0009

8 N (c) B = 3 . 9 z k 0 . 3 A.Z 24 Hi ( e ) x = 0.194 f 0.006

B = 3 . 3 f 1 . 8 A . 2 96 Hz ( m ) x = 0.368 f 0.003

z = 0.269 =I= 0.004 B = 8 .7 f. 1 . 5 A . a

( 6 ) Unpublished ACA Program No. 317 by Gantzel, Sparks, and True-

(7) G. S. Smith and L. E. Alexander, Acta Cvysl., 16, 462 (1963). (E) "International Tables for X-ray Crystallography," Vol. I, The

blood.

Kynoch Press, Birmingham, 1952.